ES2688359T3 - Adaptive central band offset filter for video encoding - Google Patents

Abstract

A video encoding method for signaling, in an encoded bit stream, the offset values used in band-based sample adaptive offset (SAO) filtering, the method comprising: grouping bands of pixel values into two or more groups of bands, wherein each band of the bands of pixel values comprises a range of pixel values, and wherein each group of the two or more groups of bands comprises one or more bands; determining offset values for the bands based on the two or more groups of bands; characterized by determining (610) center values for each of a plurality of sets of pixels, wherein for each set the center value is based on the actual values of the pixels in the set; and wherein, for each set of pixels, the grouping (620) of the pixel bands into two or more band groups is dynamically determined based on the central value for the set such that the grouping of bands into pixel sets it is adaptive.

Description

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

DESCRIPTION

Adaptive central band offset filter for video coding TECHNICAL FIELD

[1] This disclosure generally refers to video coding and more particularly refers to adaptive sample displacement (ODS) filtering in a video coding process.

BACKGROUND

[2] Digital video capabilities can be incorporated into a wide range of devices, including digital TVs, digital direct broadcasting systems, wireless broadcasting systems, personal digital assistants (PDAs), laptops or desktops, tablet computers, readers of electronic books, digital cameras, digital recording devices, digital media players, video game devices, video game consoles, cell phones or satellite radio, the so-called “smart phones”, video conferencing devices, video streaming devices and the like: Digital video devices implement video compression techniques, such as those described in the standards defined by MPEG-2, MPEG-4, ITU-T H.263, ITU-T H.264 / MPEG-4, Part 10, Advanced Video Coding (AVC), the High Efficiency Video Coding (HEVC) standard currently in disarray development and extensions of these standards. Video devices can transmit, receive, encode, decode and / or store digital video information more efficiently, implementing such video compression techniques.

[3] Video compression techniques perform spatial prediction (intra-image) and / or temporal prediction (between images) to reduce or eliminate intrinsic redundancy in video sequences. For block-based video coding, a video fragment (for example, a video frame or a part of a video frame) can be divided into video blocks, which can also be called tree blocks, encoding units (CU ) and / or coding nodes. The video blocks in an intra-encoded fragment (I) of an image are encoded by spatial prediction with respect to reference samples in contiguous blocks in the same image. Video blocks in an inter-encoded fragment (P or B) of an image may use spatial prediction with respect to reference samples in contiguous blocks in the same image, or temporal prediction with respect to reference samples in other reference images . The images can be called frames, and the reference images can be called reference frames.

[4] Spatial or temporal prediction results in a predictive block for a block to be encoded. Residual data represent pixel differences between the original block to be encoded and the predictive block. An intercoded block is encoded according to a motion vector that points to a block of reference samples that form the predictive block, and residual data indicating the difference between the encoded block and the predictive block. An intra-encoded block is encoded according to an intra-coding mode and residual data. For greater compression, residual data can be transformed from the domain of the pixels to the domain of the transformations, resulting in residual transformation coefficients, which can then be quantified. The quantified transformation coefficients, initially arranged in a two-dimensional matrix, can be scanned in order to generate a one-dimensional vector of transformation coefficients, and entropy coding can be applied to achieve even more compression.

[5] Adaptive scrolling is described in the article "CE8 Subset3: Picture Quadtree Adaptive Offset", by CM FU ET AL, 4. JCT-VC MEETING; 95. MPEG MEETING; 20-1-2011 - 28-1-2011; DAEGU; (JOINT COLLABORATION TEAM ON ISO / IEC JTC1 / SC29 / WG11 AND ITU-T SG.16 VIDEO CODING); URL:
HTTP://WFTP3.TTU.INT/AV- ARCH / JCTVC-STTE / "No. JCTVC -D122, January 15, 2011.

SUMMARY

[6] The invention is defined in the claims to which the reference is now directed. This disclosure generally describes techniques related to adaptive sample displacement (ODS) filtering and more particularly describes techniques for signaling, in a coded bit stream, the coefficients used for band-based SAO filtering. This disclosure describes techniques for adaptively determining band groupings based on a central value and using the groupings determined for the signaling offset values for the bands.

BRIEF DESCRIPTION OF THE DRAWINGS

[7]

FIG. 1 is a block diagram illustrating an example video coding and decoding system that can use the adaptive displacement filtering techniques of samples described in this

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

divulgation.

FIG. 2 is an illustration showing examples of edge displacement classifications for adaptive displacement filtering of samples based on the edge.

FIGs. 3A and 3B are illustrations showing an example of band grouping for band shift classifications for adaptive band-based sample displacement filtering.

FIG. 4 is a block diagram illustrating an example video encoder that can implement the adaptive displacement filtering techniques of samples described in this disclosure.

FIG. 5 is a block diagram illustrating an example video decoder that can implement adaptive displacement sample filtering techniques described in this disclosure.

FIG. 6 shows a flow chart illustrating the adaptive displacement filtering techniques of samples described in this disclosure.

DETAILED DESCRIPTION

[8] Adaptive Sample Displacement Filtering (ODS) is a type of loop filtering used in video coding. In general, adding offset values to pixels in a video frame (for example, a reconstructed image) can in some cases improve the encoding without greatly increasing the bit overhead required to store or transmit encoded video data. The improvement in coding that is potentially obtained as a result of the ODS filtering may be, for example, that a decoded image is more similar to an original image. As will be explained in greater detail below, ODS techniques allow different offset values to be applied to different pixels (or blocks of pixels) depending on the pixel classification metrics (or blocks), such as edge metrics, metrics of band or other types of metrics.

[9] As will be explained in more detail below, in some configurations, an SAO filter unit may be configured to perform two types of displacement filtering, generally referred to in this disclosure as bandwidth filtering and edge shifting filtering. An SAO filter unit may also sometimes not apply displacement, which can be considered a third type of displacement filtering. The type of displacement filtering applied by an SAO filter can be explicitly or implicitly signaled to a video decoder. By applying edge shift filtering, pixels can be classified based on the edge information of a coding unit, and a offset for pixels can be determined based on the edge classification. As will be explained in greater detail below with reference to FIG. 2, there are typically four variations of ODS based on the edge, where the value of a pixel is compared with two of its eight neighboring pixels. The two pixels that are used to compare depend on the variation of the offset based on the edge that is used. Based on the difference in magnitude, a shift is added to the pixel value.

[10] When applying band offset filtering, pixels can be classified into different bands based on a pixel value, such as an intensity value, with each band having an associated offset. A band includes a range of pixel values. For example, pixel values ranging from 0 to 255 can be divided into 32 equal bands (labeled 0 to 31), so that pixel values 0-7 are a first band, pixel values 8-15 they are a second band, pixel values 16-23 are a third band, and so on for the thirty-two bands. Bands can be used to determine which particular offset value to apply to a pixel or group of pixels. For example, if a pixel has a value of 10 (which is within the second band, that is, values 8-15, in the previous example), then a offset associated with the second band can be added to the pixel value.

[11] For the purpose of signaling and generating displacements for the various bands, the bands can be grouped into two or more groups. For bandwidth filtering, pixels can, for example, be classified into thirty-two bands (bands 0-31) as described above, and the bands can be grouped into two groups (for example, two groups of sixteen bands, a group of four bands and a group of twenty-eight bands, a group of eight bands and a group of twenty-four bands or other similar groupings). Band groupings can be used to determine the order in which the offset values for the bands are signaled in the encoded video bit stream, and / or can be used to determine if a particular band has a offset value other than zero. The displacements for the bands can be signaled using differential coding techniques in which a current value is signaled as a difference between the current value and a previous value. The amount of bit savings achieved through the use of such coding techniques may in some cases depend on the order in which the values are indicated. For example, greater bit savings can be achieved in cases where offset values that are close to their value are signaled consecutively. In some cases, it can be assumed that all offset values for a particular group, such as the second group, are zero.

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

[12] According to some techniques, band groupings are static. For example, according to one technique, the middle sixteen bands (bands 8-23) always constitute a group of sixteen bands, while the eight leftmost bands (bands 0-7) and the eight most bands the right (bands 24-31) constitute a second group of sixteen bands. This disclosure describes techniques for dynamically determining band groupings. As will be explained in greater detail below, a video encoder or video decoder can dynamically determine a band grouping by dynamically determining a central value based on pixel values. In some configurations, an indication of a central value in the bit stream of encoded video may be signaled so that a video decoder does not need to determine the core value based on pixel values, but can determine the core value based on the explicit signaling in the bit stream. The central value can be used to determine a first group of bands, and the remaining bands can constitute a second group of bands. As will be clear from the examples below, the term "core value" as used in this disclosure generally refers to any value that can be used to align or approximately align the center of a group of bands with a particular pixel value. Therefore, a central value as used in this disclosure may be the particular pixel value on which the center of the band group is aligned or it may be some other value, such as an initial or final pixel value for a band or band group. In some cases, instead of identifying a pixel value, a central value can identify a particular band such as the first band of a group, the last band of a group or some other identification on which the center of a group can be aligned. of bands at a particular pixel value.

[13] The dynamic determination of band groupings can, in some cases, improve the signaling of the displacement coefficients for the ODS filtering and, therefore, can improve the overall quality of the video coding. This dynamic determination of band groupings can also be used to determine for which band groups the SAO filtering will benefit the video coding quality most, assuming that other bands have a zero offset value.

[14] FIG. 1 is a block diagram illustrating an example 10 video encoding and decoding system that can use the adaptive ODS techniques described in this disclosure. As shown in FIG. 1, the system 10 includes a source device 12 that generates encoded video data that a destination device 14 will decode at a later time. The source device 12 and the destination device 14 can comprise any of a wide range of devices, including desktop computers, folding computers (ie laptops), tablet computers, decoders, telephone equipment such as so-called "smart phones" ”, The so-called“ smart ”panels, televisions, cameras, display devices, digital media players, video game consoles, a video streaming device or the like. In some cases, the source device 12 and the destination device 14 may be equipped for wireless communication.

[15] The target device 14 can receive the encoded video data to be decoded, via a link 16. The link 16 may comprise any type of media or device capable of moving the encoded video data from the source device 12 to the destination device 14. In one example, the link 16 may comprise a communication means to allow the source device 12 to transmit the encoded video data directly to the destination device 14 in real time. The encoded video data may be modulated according to a communication standard, such as a wireless communication protocol, and transmitted to the destination device 14. The communication medium may comprise any wireless or wired communication means, such as a spectrum of radio frequency (RF) or one or more physical transmission lines. The communication medium may be part of a packet-based network, such as a local area network, a wide area network or a global network such as the Internet. The communication medium may include routers, switches, base stations or any other equipment that may be useful to facilitate communication from the source device 12 to the destination device 14.

[16] Alternatively, the encoded data can be output from the output interface 22 to a storage device 32. Similarly, the encoded data can be accessed from the storage device 32 via an input interface. The storage device 32 may include any of a variety of data storage, distributed or local access means, such as a hard disk drive, Blu-ray discs, DVD discs, CD-ROM discs, flash memory, volatile or non-volatile memory or any other suitable digital storage medium, for storing encoded video data. In a further example, the storage device 32 may correspond to a file server or other intermediate storage device that can retain the encoded video generated by the source device 12. The destination device 14 can access the stored video data from the storage device 32 through continuous transmission or download. The file server can be any type of server capable of storing encoded video data and transmitting that encoded video data to the target device 14. Examples of file servers include a web server (for example, for a website), a FTP server, network connected storage devices (NAS) or a local disk drive. The destination device 14 can access the encoded video data through any standard data connection, including an Internet connection. This may include a wireless channel (for example, a Wi-Fi connection), a wired connection (for example, DSL, cable modem, etc.), or a combination of both

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

suitable for accessing encoded video data stored on a file server. The transmission of encoded video data from the storage device 32 may be a continuous transmission, a download transmission or a combination of both.

[17] The techniques of this disclosure are not necessarily limited to wireless applications or configurations. The techniques can be applied to video coding, in support of any of a variety of multimedia applications, such as television broadcasts over the air, cable television transmissions, satellite television transmissions, continuous video transmissions, by for example, via the Internet, digital video coding for storage in a data storage medium, digital video decoding stored in a data storage medium, or other applications. In some examples, the system 10 may be configured to support unidirectional or bidirectional video transmission to support applications such as continuous video transmission, video playback, video broadcasting and / or video telephony.

[18] In the example of FIG. 1, the source device 12 includes a video source 18, a video encoder 20 and an output interface 22. In some cases, the output interface 22 may include a modulator / demodulator (modem) and / or a transmitter. In the source device 12, the video source 18 may include a source such as a video capture device, for example, a camcorder, a video file containing previously captured video, a video feed interface for receiving video from a video content provider and / or a computer graphics system to generate computer graphics data such as the source video, or a combination of such sources. As an example, if the video source 18 is a camcorder, the source device 12 and the destination device 14 can form the so-called camera phones or videophone. However, the techniques described in this disclosure may be applicable to video coding in general, and may be applied to wireless and / or wired applications.

[19] The video captured, pre-captured or computer generated can be encoded by the video encoder 20. The encoded video data can be transmitted directly to the destination device 14 via the output interface 22 of the source device 12. The video data encoded also (or alternatively) may be stored in the storage device 32 for later access by the destination device 14 or other devices, for decoding and / or playback.

[20] The target device 14 includes an input interface 28, a video decoder 30 and a display device 32. In some cases, the input interface 28 may include a receiver and / or a modem. The input interface 28 of the destination device 14 receives the video data encoded by the link 16. The video data encoded, communicated by the link 16, or provided on the storage device 32, may include a variety of generated syntactic elements. by video encoder 20, for use by a video decoder, such as video decoder 30, in decoding video data. Said syntactic elements may be included with the encoded video data, transmitted in a communication medium, stored in a storage medium or stored in a file server.

[21] The display device 32 may be integrated with, or be external to, the target device 14. In some examples, the target device 14 may include an integrated display device and also be configured to interconnect with a display device external. In other examples, the destination device 14 may be a display device. In general, the display device 32 displays the decoded video data before a user and can comprise any of a variety of display devices, such as a liquid crystal display (LCD), a plasma screen, an organic diode screen light emitters (OLED) or other type of display device.

[22] The video encoder 20 and the video decoder 30 can operate in accordance with a video compression standard, such as the High Efficiency Video Coding (HEVC) standard, currently being developed, and can be shaped to the HEVC Test Model (HM). A recent draft of the HEVC standard, called “Working Draft 8 of the HEVC” or “WD8”, is described in document JCTVC-H1003, by Bross et al., Entitled "High Efficiency Video Coding (HEVC) Text Specification Draft 8 "[" Textual descriptive memory of High Efficiency Video Coding (HEVC), Draft 8 "], Joint Collaboration Team on Video Coding (JCT-VC) of ITU-T SG16 WP3 and ISO / IEC JTC1 / SC29 / WG11, 10th meeting: Stockholm, SE, July 11-20, 2012, which, as of October 2, 2012, can be downloaded from the following link:
http: //phenix.int- evry.fr/jct/doc_end_user/documents/10_Stockholm/wg11JCTVC-J1003-v8.zip.

[23] Alternatively, video encoder 20 and video decoder 30 can operate in accordance with other private or industrial standards, such as ITU-T H.264, alternatively called MPEG-4, Part 10 , Advanced Video Coding (AVC), or extensions of these standards. However, the techniques of this disclosure are not limited to any particular coding standard. Other examples of video compression standards include MPEG-2 and ITU-T H.263.

[24] Although not shown in FIG. 1, in some aspects, the video encoder 20 and the video decoder 30 may each be integrated in an audio encoder and decoder, and may include

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

Suitable MUX-DEMUX units, or other hardware and software, to manage both audio and video encoding in a common data stream or in different data streams. If applicable, in some examples, MUX-DEMUX units can conform to the ITU H.223 multiplexing protocol or other protocols, such as the user datagram protocol (UDP).

[25] Video encoder 20 and video decoder 30 can be implemented as any of a variety of suitable encoder circuits, such as one or more microprocessors, digital signal processors (DSP), application-specific integrated circuits ( ASIC), field programmable door matrices (FPGA), discrete logic, software, hardware, firmware or any combination thereof. When the techniques are partially implemented in software, a device may store instructions for the software in a suitable non-transient, computer-readable medium, and execute the instructions in hardware by one or more processors to perform the techniques of this disclosure. Each of the video encoder 20 and the video decoder 30 may be included in one or more encoders or decoders, any of which may be integrated as part of a combined encoder / decoder (CODEC) in a respective device.

[26] The JCT-VC team is working on the development of the HEVC standard. HEVC's standardization efforts are based on an evolving model of a video coding device called the HEVC (HM) test model. The HM assumes several additional capabilities of video encoding devices with respect to existing devices according, for example, to the ITU-T H.264 / AVC standard. For example, while the H.264 standard provides nine intra-prediction coding modes, the HM can provide up to thirty-three intra-prediction coding modes.

[27] In general, the HM's operating model describes that a video frame or image can be divided into a sequence of wooded blocks or maximum size coding units (LCUs), which include both luma and chroma samples. A tree block has an end similar to that of a macroblock of the H.264 standard. A fragment includes a certain number of consecutive wooded blocks in coding order. A plot or video image can be divided into one or more fragments. Each wooded block can be divided into coding units (CU) according to a quad tree. For example, a tree block, such as a root node of the quaternary tree, can be divided into four child nodes, and each child node can, in turn, be a parent node and be divided into four other child nodes. A final, undivided child node, as the leaf node of the quaternary tree, comprises an encoding node, that is, a block of encoded video. The syntactic data associated with an encoded bit stream can define a maximum number of times a tree block can be divided, and they can also define a minimum size of the coding nodes.

[28] A CU includes a coding node and prediction units (PU) and transformation units (TU) associated with the coding node. A CU size corresponds to a coding node size and must be square. The size of the CU can vary from 8x8 pixels to the size of the wooded block, with a maximum of 64x64 pixels or more. Each CU may contain one or more PU and one or more TU. The syntactic data associated with a CU may describe, for example, the division of the CU into one or more PU. The division modes may differ based on whether the CU is encoded in skip or direct mode, encoded in intra-prediction mode or encoded in inter-prediction mode. PUs can be divided to have a non-square shape. The syntactic data associated with a CU can also describe, for example, the division of the CU into one or more TU according to a quad tree. A TU can have a square or non-square shape.

[29] The HEVC standard admits transformations according to TU, which may be different for different Cu. The size of the TU is typically based on the size of the PUs of a given CU, defined for a divided LCU, although this may not always be the case. TUs are typically the same size or smaller than PU. In some examples, the residual samples corresponding to a CU can be subdivided into smaller units by a quadruple tree structure known as "residual quadruple tree" (RQT). The leaf nodes of the RQT can be called transformation units (TU). Pixel difference values, associated with TU, can be transformed to generate transformation coefficients, which can be quantified.

[30] In general, a PU includes data related to the prediction process. For example, when the PU is encoded intramodally, the PU may include data describing an intraprediction mode for the PU. As another example, when the PU is intercoded, the PU may include data defining a motion vector for the PU. The data defining the motion vector for a PU can describe, for example, a horizontal component of the motion vector, a vertical component of the motion vector, a resolution for the motion vector (for example, precision of a quarter pixel or accuracy of an eighth pixel), a reference image pointed to by the motion vector and / or a list of reference images (for example, List 0, List 1 or List C) for the vector of movement.

[31] In general, a TU is used for transformation and quantification processes. A given CU that has one or more PUs can also include one or more transformation units (TU). After prediction, the video encoder 20 can calculate the residual values corresponding to the PU. Residual values

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

they include values of pixel differences that can be transformed into transformation coefficients, quantified and scanned by the TU, to generate series transformation coefficients for entropy coding. This disclosure typically uses the term "video block" to refer to a coding node of a CU. In some specific cases, this disclosure may also use the term "video block" to refer to a tree block, that is, an LCU or a CU, which includes an encoding node and PU and TU.

[32] A video sequence typically includes a series of frames or video images. A group of images (GOP) generally comprises a series of one or more of the video images. A GOP may include syntactic data in a GOP header, in a header of one or more of the images or in other locations, which describe a certain number of images included in the GOP. Each fragment of an image may include syntactic fragment data that describes a coding mode for the respective fragment. A video encoder 20 typically acts on video blocks within individual video fragments in order to encode the video data. A video block may correspond to an encoding node within a CU. Video blocks may have fixed or variable sizes and may differ in size according to a specified coding standard.

[33] In one example, the HM supports prediction in various PU sizes. Assuming that the size of a particular Cu is 2Nx2N, the HM supports intra-prediction in PU sizes of 2Nx2N or NxN and inter-prediction in symmetric PU sizes of 2Nx2N, 2NxN, Nx2N or NxN. The HM also supports asymmetric division for interprediction in PU sizes of 2NxnU, 2NxnD, nLx2N and nRx2N. In asymmetric division, one direction of a CU is not divided, while the other direction is divided into 25% and 75%. The part of the CU corresponding to the 25% division is indicated by an “n” followed by an indication of “up”, “down”, “left” or “right”. Thus, for example, "2NxnU" refers to a CU of size 2Nx2N that is horizontally divided with a PU of size 2Nx0.5N above and a PU of size 2Nx1.5N below.

[34] In this disclosure, "NxN" and "N by N" can be used interchangeably to refer to the pixel dimensions of a video block in terms of vertical and horizontal dimensions, for example, 16x16 pixels or 16 by 16 pixels. . In general, a block of size 16x16 will have 16 pixels in the vertical direction (y = 16) and 16 pixels in the horizontal direction (x = 16). Likewise, a block of size NxN generally has N pixels in a vertical direction and N pixels in a horizontal direction, where N represents a non-negative integer value. Pixels in a block can be arranged in rows and columns. In addition, it is not necessary that the blocks necessarily have the same number of pixels in the horizontal direction and in the vertical direction. For example, the blocks can comprise NxM pixels, where M is not necessarily equal to N.

[35] Following the coding of intra-prediction or inter-prediction by means of the PUs of a CU, the video encoder 20 can calculate residual data for the TUs of the CU. PUs can comprise pixel data in the spatial domain (also called pixel domain) and TUs can comprise coefficients in the transformation domain after the application of a transformation, for example, a discrete cosine transformation (DCT), a integer transformation, a waveform transformation or a similar transformation from a conceptual point of view to residual video data. The residual data may correspond to pixel differences between pixels of the uncoded image and the prediction values corresponding to the PU. The video encoder 20 can form the TUs including residual data for the CU and then transform the TUs to generate transformation coefficients for the CU.

[36] After any transformation to generate transformation coefficients, the video encoder 20 can quantify the transformation coefficients. Quantification refers, in general, to a process in which the transformation coefficients are quantified to possibly reduce the amount of data used to represent the coefficients, providing additional compression. The quantification process can reduce the bit depth associated with some or all of the coefficients. For example, a value of n bits can be rounded down to a value of m bits during quantization, where n is greater than m.

[37] In some examples, video encoder 20 can use a predefined scan order to scan quantified transformation coefficients, to produce a serial vector that can be encoded by entropy. In other examples, video encoder 20 can perform adaptive scanning. After scanning the quantified transformation coefficients to form a one-dimensional vector, the video encoder 20 can perform entropy coding of the one-dimensional vector, for example, according to the adaptive variable length coding according to the context (CAVLC), adaptive binary arithmetic coding according to the context (CABAC), adaptive binary arithmetic coding according to the context based on syntax (SBAC), entropy coding by division of probability intervals (PIPE) or other entropy coding procedures. The video encoder 20 can also perform the entropy coding of the syntactic elements associated with the encoded video data, for use by the video decoder 30 in decoding the video data.

[38] To perform the CABAC, video encoder 20 can assign a context within a contextual model to a symbol to be transmitted. The context may refer, for example, to whether the contiguous values of the symbol are nonzero or not. To perform the CAVLC, video encoder 20 can select a

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

variable length code for a symbol to be transmitted. The code words in the VLC can be constructed in such a way that the relatively shorter codes correspond to more likely symbols, while the longer codes correspond to less probable symbols. In this way, the use of the VLC can achieve savings in bits with respect, for example, to the use of code words of equal length for each symbol to be transmitted. The determination of the probability can be based on a context assigned to the symbol.

[39] In general, adding offset values to pixels in a video frame may improve coding in some cases. For example, offset values can be applied to the pixels of a reconstructed video block to compensate for lighting changes, quantization errors, or more in general, to make decoded video data more like data from original video ODS techniques allow different offset values to be applied to different pixels (or blocks of pixels) depending on the pixel values of a pixel (or block). The offset value that will be applied to a pixel can be determined based on the value of a pixel. For example, if a pixel has a value that is within a first band, then a offset associated with the first band can be applied to the pixel. If the pixel has a value that is within a second band, then a offset associated with the second band can be applied to the pixel, and so on for all bands.

[40] In one type of ODS implementation, each partition (consisting of a set of LCUs) can have one of three types of offset (also called pixel classifications). The three types of displacement are not displacement, type of displacement based on the 0/1 band classification and type based on the 0/1/2/3 edge classification. Each type of band classification offset has 16 possible offset values, while each type based on edge classification has 4 possible offset values. If one of these types of offset is chosen to be used for the partition, the information indicating the corresponding type of offset and offset values can be signaled in the bit stream of encoded video.

[41] FIG. 2 is a conceptual diagram showing the four possible edge offset classifications used in an SAO filtering implementation. In the example of FIG. 2, the type of border offset classifies each pixel based on the edge information. For each of the edge classifications shown in FIG. 2, a border type is calculated for the current pixel (i.e. a particular pixel that is being encoded) by comparing the value of the current pixel (C) with the values of the neighboring pixels (1 and 2). For the SAO edge offset of the zero classification (SAO_EO_0), the current pixel is compared with the neighboring left and right pixels. For the SAO edge offset of classification one (SAO_EO_1), the current pixel is compared to the upper and lower neighboring pixels. For the SAO edge offset of classification two (SAO_ EO_2), the current pixel is compared to the upper left and lower right neighboring pixels. For the SAO edge offset of classification three (SAO_EO_3), the current pixel is compared with the lower left and upper right neighboring pixels.

[42] Initially, the current pixel border type is assumed to be zero. If the value of the current pixel C is equal to the values of the neighboring left and right pixels (1 and 2), the border type remains zero. If the value of the current pixel C is greater than the value of the neighboring pixel 1, the border type is increased by one. If the value of the current pixel C is less than the value of the neighboring pixel 1, the border type is reduced by one. Similarly, if the value of the current pixel C is less than the value of neighboring pixel 2, the border type is increased by one, and if the value of the current pixel C is less than the value of neighboring pixel 2, the type Edge is reduced by 1.

[43] As such, the current pixel V can have a border type of -2, -1, 0, 1 or 2. The border type is -2 if the value of the current pixel C is less than both values of the neighboring pixels 1 and 2. The border type is -1 if the value of the current pixel C is less than a neighboring pixel, but equal to the other neighboring pixel. The border type is 0 if the value of the current pixel C is the same as that of both neighboring pixels, or if the value of the current pixel C is greater than a neighboring pixel, but less than that of the other neighboring pixel. The border type is 1 if the value of the current pixel C is greater than a neighboring pixel, but equal to the other neighboring pixel. The border type is 2 if the value of the current pixel C is greater than both values of neighboring pixels 1 and 2. For each non-zero edge type value, four offset values are determined and indicated in the bit stream video encoded for use by a decoder (ie, eoffset .2, eoffset .1, eoffset 1, eoffset 2).

[44] In view of the above description, for each edge offset classification, edge type values can be calculated with the following pseudocode:

EdgeType = 0;

if (C> Pixel1) EdgeType = EdgeType + 1;

if (c <Pixel1) EdgeType = EdgeType - 1;

if (c> Pixel2) EdgeType = EdgeType + 1;

if (c <Pixel2) EdgeType = EdgeType - 1

[45] FIG. 3A is a conceptual diagram showing sample bands based on intensity values.

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

As an example, assume that pixel values range from 0 to 255, although other ranges can also be used. In such an example, the maximum value shown in FIG. 3A would be equal to 255, and each of the thirty-two bands shown in FIG. 3A would have a range of 8. The leftmost band would be pixel values 0-7, the next band would be for pixel values of 8-15, the next band would be for pixel values 16-23, and so on, to the rightmost band that would be for pixel values 248-255. For band offset, pixels are classified into different bands based on intensity (that is, for band offset classification, pixels are classified into one of the thirty-two bands). Based on the band in which the value of a pixel is found, a shift is added to the pixel. For example, if a pixel has a value of 19, then the pixel value is within the third band that varies from the pixel value 16 to 23. Therefore, a offset associated with the third band would be added to the value 19 pixel

[46] For signaling purposes of the offset values associated with each band, the bands can be grouped into two or more groups. In some implementations, the sixteen bands in the center (bands 823) are classified in one group and the remaining bands (bands 0-7 and 24-31) are classified in a second group. For each group of bands, 16 offset values (ie, boffsetü, ..., boffset15) are determined and signaled in the video bit stream encoded for use by a video decoder. In some implementations, it can be assumed that all offset values for a group, such as the second group, are 0, in which case signaling of offset values for that group should not be included in the encoded video bit stream.

[47] As shown in FIG. 3A, the sixteen middle bands (bands 8-23) constitute the first group of bands, while the eight most left bands (band 0-7) constitute the first part of a second group of bands, and the eight more bands on the right (bands 24-31) constitute a second part of the second group of bands. According to current techniques, this band grouping is fixed. This disclosure describes techniques for adaptively determining band groupings. For example, this disclosure describes techniques for determining a central value for a set of pixels based on pixel values. The central value can, for example, be determined based on a given center for a band group, and the center for the band group can be determined based on an average value for the set of pixels, a median value of a set of pixels. , or determining the pixel where the peak of a histogram is located.

[48] Pixel value bands can be divided into groups based on the central value. For example, the central value can be used to determine a center for a group of bands. Again assuming thirty-two bands (bands 0-31) and two groups, the eight bands to the right of the center and the eight bands to the left of the center may be the first group of bands. The remaining sixteen bands may be the second group. Therefore, unlike previous techniques in which a group always consisted of the sixteen middle bands, this disclosure describes techniques in which band groupings can be adaptive based on a core value that is determined based on the values of real pixels

[49] FIG. 3B shows one of many possible groupings that can be used in accordance with the techniques of this disclosure. In the example of FIG. 3B, a center is determined near band 18. Consequently, the first group of bands includes the eight bands to the left of the center (bands 11-18) and the eight bands to the right of the center (bands 19-26) . The second group of bands includes the remaining bands, which in this example are bands 0-4 and bands 27-31. Again, FIG. 3B represents one of the many possible scenarios. The main objective of FIG. 3B is to show that, according to the techniques of this disclosure, the center for determining the first group of bands can be determined adaptively and does not need to be fixed at the location shown in FIG. 3A or fixed at any other location. The techniques for determining a central value and adaptively determining band groupings will be explained in more detail below.

[50] FIG. 4 is a block diagram illustrating an example video encoder 20 that can implement the SAO filtering techniques described in this disclosure. The video encoder 20 can perform intra-coding and inter-coding of video blocks within video fragments. Intra-coding relies on spatial prediction to reduce or eliminate spatial redundancy in the video within a given frame or video image. Intercoding is based on temporal prediction to reduce or eliminate temporal redundancy in the video within frames or adjacent images of a video sequence. The intra mode (mode I) may refer to any of several spatial compression modes. Inter modes, such as unidirectional prediction (P mode) or bi-prediction (B mode), can refer to any of several time-based compression modes.

[51] In the example of FIG. 4, the video encoder 20 includes a partition module 35, a prediction module 41, a reference image memory 64, an adder 50, a transformation module 52, a quantization module 54 and an entropy coding module 56. The prediction module 41 includes a motion estimation module 42, a motion compensation module 44 and an intra-prediction module 46. For the reconstruction of video blocks, the video encoder 20 further includes a video module. reverse quantization 58, a reverse transformation module 60 and an adder 62. The unblocking filter 72 can also be included to filter the block boundaries to eliminate blockage distortions of the video.

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

rebuilt. As shown in FIG. 4, the video encoder 20 also includes additional loop filters, including the adaptive sample shift filter (SAO) 74 and the adaptive loop filter (ALF) 76. Although the unlock filter 72, the SAO filter 74 and the ALF 76 are shown as loop filters in FIG. 4, in some configurations, the release filter 72, the SAO filter 74 and the ALF 76 can be implemented as post-loop filters.

[52] As depicted in FIG. 4, the video encoder 20 receives video data, and the division module 35 divides the data into video blocks. This division may also include the division into fragments, elements or other larger units, as well as the division of video blocks, for example, according to a quaternary tree structure of some LCU and CU. The video encoder 20 illustrates, in general, the components encoding video blocks of a video fragment to be encoded. The fragment can be divided into several video blocks (and possibly into sets of video blocks called elements). The prediction module 41 may select one from a plurality of possible coding modes, such as one from a plurality of intra-coding modes, or one from a plurality of inter-coding modes, for the current video block, based in error results (for example, coding speed and distortion level). The prediction module 41 can provide the resulting intra-encoded or inter-encoded block to adder 50 to generate residual block data, and to adder 62 to reconstruct the encoded block for use as a reference image.

[53] The intra-prediction module 46, within the prediction processing module 41, can perform the intra-predictive coding of the current video block with respect to one or more neighboring blocks in the same frame or fragment as the block that is going to be encoded, to provide spatial compression. The motion estimation module 42 and the motion compensation module 44, within the prediction module 41, perform the inter-predictive coding of the current video block with respect to one or more predictive blocks in one or more reference images, to provide temporary compression.

[54] The motion estimation module 42 may be configured to determine the mode of interpretation for a video fragment according to a predetermined pattern for a video sequence. The predetermined pattern may designate video fragments in the sequence as predicted fragments (P fragments), predicted fragments in two directions (sectors B), or generalized P / B fragments (GPB fragments). The motion estimation module 42 and the motion compensation module 44 may be highly integrated, but are illustrated separately for conceptual purposes. Motion estimation, performed by motion estimation module 42, is the process of generating motion vectors, which estimate movement for video blocks. A motion vector, for example, may indicate the displacement of a PU from a video block of a current video frame or image with respect to a predictive block of a reference image.

[55] A predictive block is a block that is found to correspond closely with the PU of the video block to be encoded in terms of pixel difference, which can be determined by the sum of an absolute difference (SAD), sum of squared difference (SSD) or other difference metrics. In some examples, video encoder 20 can calculate values for fractional pixel positions of reference images stored in reference image memory 64. For example, video encoder 20 can interpolate values of positions of a quarter pixel, positions of an eighth pixel or other fractional pixel positions of the reference image. Therefore, the motion estimation module 42 can perform a motion search with respect to the positions of full pixels and the positions of pixel fractions, and emit a motion vector with an accuracy of pixel fractions.

[56] Motion estimation module 42 calculates a motion vector for a PU of a video block in an inter-encoded fragment, comparing the position of the PU with the position of a predictive block of a reference image. The reference image may be selected from a first list of reference images (List 0) or a second list of reference images (List 1), each of which identifies one or more reference images stored in the image memory of reference 64. The motion estimation module 42 sends the calculated motion vector to the entropy coding module 56 and the motion compensation module 44.

[57] Motion compensation, performed by the motion compensation module 44, may involve extracting or generating the predictive block based on the motion vector determined by the motion estimate, possibly performing interpolations up to the sub-pixel accuracy. After receiving the motion vector for the PU of the current video block, the motion compensation module 44 can locate the predictive block to which the motion vector points in one of the reference image lists. The video encoder 20 forms a residual video block by subtracting the pixel values of the predictive block from the pixel values of the current video block being encoded, generating pixel difference values. Pixel difference values form residual data for the block, and may include components of luminance and chroma difference. Adder 50 represents the component or components that perform this subtraction operation. The motion compensation module 44 can also generate syntactic elements associated with the video blocks and the video fragment for use by the video decoder 30 in decoding

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

of the video blocks of the video fragment.

[58] The intra-prediction module can perform the intra-prediction in a current block, as an alternative to the inter-prediction performed by the motion estimation module 42 and the motion compensation module 44, as described previously. In particular, the intra-prediction module 46 can determine an intra-prediction mode to be used to encode a current block. In some examples, the intra-prediction module 46 may encode a current block using several intra-prediction modes, for example, during independent coding passes, and the intra-prediction module 46 (or the mode selection module 40 , in some examples) you can select a suitable intra-prediction mode to use, among the modes tested. For example, the intra-prediction module 46 can calculate velocity distortion values using a velocity distortion analysis for the various intra-prediction modes tested, and select the intra-prediction mode that has the best distortion characteristics of Speed between tested modes. The speed distortion analysis determines, in general, a magnitude of distortion (or error) between an encoded block and an original uncoded block that was encoded to generate the encoded block, as well as a bit rate (i.e., a number of bits) used to generate the encoded block. The intra-prediction module 46 can calculate ratios from the distortions and speeds for the various encoded blocks, to determine which intra-prediction mode has the best speed distortion value for the block.

[59] In any case, after selecting an intra-prediction mode for a block, the intra-prediction module 46 can provide information indicating the intra-prediction mode selected for the block to the entropy coding module 56. The Entropy coding module 56 may encode the information indicating the selected intra-prediction mode in accordance with the techniques of this disclosure. The video encoder 20 may include configuration data in the transmitted bit stream, which may include a plurality of index tables of intra-prediction modes and a plurality of index tables of modified intra-prediction modes (also called tables code word assignment), coding context definitions for several blocks and indications of the most likely intra-prediction mode, an index table of intra-prediction modes and a table of indexes of modified intra-prediction modes to use for each of the contexts.

[60] After the prediction module 41 generates the predictive block for the current video block, either by inter-prediction or intra-prediction, the video encoder 20 forms a residual video block by subtracting the predictive block to the current video block. The residual video data in the residual block may be included in one or more TU and applied to the transformation processing module 52. The transformation module 52 transforms the residual video data into residual transformation coefficients, using a transformation, such as a discrete cosine transformation (DCT) or a conceptually similar transformation. The transformation processing module 52 can convert the residual video data, from a pixel domain to a transformation domain, such as a frequency domain.

[61] The transformation module 52 can send the resulting transformation coefficients to the quantization module 54. The quantization module 54 quantifies the transformation coefficients to further reduce the bit rate. The quantification process can reduce the bit depth associated with some or all of the coefficients. The degree of quantification can be modified by adjusting a quantification parameter. In some examples, the quantization module 54 can then perform a matrix scan that includes the quantified transformation coefficients. Alternatively, the entropy coding module 56 can perform the scan.

[62] After quantification, the entropy coding module 56 encodes entropy the coefficients of

quantified transformation. For example, the entropy coding module 56 can perform adaptive variable length coding according to the context (CAVLC), an adaptive binary arithmetic coding according to the context (CABAC), an adaptive binary arithmetic coding according to the context based on the syntax ( SBAC), an entropy coding by division of probability intervals (PIPE) or other entropy coding methodology or technique. Following the entropy coding performed by the entropy coding module 56, the encoded bit stream can be transmitted to the video decoder 30, or archived for later transmission or recovery by the video decoder 30. The entropy coding module 56 also can

perform entropy coding of motion vectors and other syntactic elements for the

Current video snippet that is being encoded.

[63] The inverse quantization module 58 and the inverse transformation module 60 apply a quantification

inverse and an inverse transformation, respectively, to reconstruct the residual block in the pixel domain, for later use as a reference block of a reference image. The motion compensation module 44 can calculate a reference block by adding the residual block to a predictive block of one of the reference images of one of the reference image lists. The compensation module of

movement 44 can also apply one or more interpolation filters to the reconstructed residual block to calculate

fractional pixel values for use in motion estimation. Adder 62 adds the reconstructed residual block to the predictive block with motion compensation generated by the compensation module of

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

movement 44 to generate a reference block for storage in the reference image memory 64.

[64] Prior to storage in memory 64, the reconstructed residual block can be filtered by one or more filters. If desired, the release filter 72 can also be applied to filter the reconstructed residual blocks in order to eliminate blockage distortions. Other loop filters can also be used (either in the encoding loop or after the encoding loop) to smooth pixel transitions or otherwise improve video quality. An example of these other filters is the SAO filter 74. The motion estimation module 42 and the motion compensation module 44 can use the reference block as a reference block to perform the inter-prediction of a block in an image or subsequent video frame.

[65] The SAO 74 filter can determine offset values for the SAO filtering in a manner that improves the quality of the video encoding. Improving the quality of video encoding may, for example, involve the determination of offset values that make a reconstructed image more closely resemble an original image. The video encoder 20 can, for example, encode the video data using multiple passes with different offset values and choose, for inclusion in an encoded bit stream, the offset values that offer the best encoding quality, determined based on in a speed distortion calculation.

[66] In some configurations, the SAO 74 filter can be configured to apply two types of offset (eg, band offset and edge offset) as described above. The SAO filter 74 may also sometimes not apply displacement, which can be considered a third type of displacement. The type of offset applied by the SAO 74 filter can be explicitly or implicitly signaled to a video decoder. When applying the offset of the border, the pixels can be classified according to the border information according to FIG. 2 and a filter can be determined based on the classification of the edges. By applying the band offset, the SAO 74 filter can classify pixels in different bands based on a pixel value, such as an intensity value, each band having an associated offset.

[67] In order to signal and generate the displacements for the various bands, the SAO filter 74 can group the thirty-two bands into two or more groups, as shown in the examples of FIGs. 3A and 3B. The SAO 74 filter can, for example, group each of the thirty-two bands into two groups (ie, two groups of sixteen bands). Although this disclosure generally describes grouping bands in two groups of equal size, more than two groups and groups of unequal sizes can also be used. This disclosure describes techniques to determine how these groups are determined.

[68] Each band can cover a range of pixel values. A pixel value can be, for example, a pixel intensity value and can describe any of an RGB value, a luminance value, a chrominance value, or any other type of pixel value. The techniques described in this disclosure may, for example, be applied to chroma samples of video data, luma samples of video data, or chroma and luma samples of video data.

[69] In some implementations, a first group of bands may focus on the center of the range of possible pixel values (ie, max / 2). Thus, for example, if the range of pixel values is 0-255, the center of the first band group could be between pixel value 127 and pixel value 128, with the first group encompassing values of 64-191. A first part of the second group can cover pixel values 0-63 and a second part of the second group can cover pixel values from 192 to 255. The first group includes sixteen bands, and the second group includes sixteen bands (eight bands in the first part and eight bands in the second part). This grouping example is shown in FIG. 3A. While the examples of this disclosure generally assume that each band of the thirty-two bands has the same size, in some implementations, bands of unequal size can be used.

[70] In accordance with the techniques of this disclosure, the SAO 74 filter can adaptively determine the center of the first group instead of using a static band grouping. The SAO filter 74 may, for example, determine the center of the first group based on a mean or median pixel value for the pixels in a region of a frame. The sAo 74 filter can also determine the center of the first group based on the pixel where the peak of a histogram is located. Once the SAO filter 74 determines the center of the first group, the SAO filter 74 can then determine the second group. Again, assuming thirty-two bands, the first group may include approximately eight bands to the left of the center and approximately eight bands to the right of the center. The remaining bands can form the second group.

[71] As an example, suppose the maximum pixel value is 256, and the SAO filter 74 determines a center for a first band group, based on any of the techniques described above, at the pixel value 90. In that case , in relation to the static grouping of FIG. 3A, the first group of bands can move to the left, making the first part of the second group smaller, while increasing the second part of the second group. However, if the SAO filter 74 determines a center for the first

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

band group at 170, then the first group moves to the right in relation to the static grouping of FIG. 3A, making the first part of the second group larger and the second part of the second group smaller. The SAO filter 74 may determine the center according to any of the techniques described in this disclosure to determine a mean, median or peak of a histogram value, or it may be determined in some other way.

[72] After determining a center, the SAO filter 74 can determine a first group of bands. The SaO 74 filter can determine the first group of bands by identifying the eight bands to the left of the center and the eight bands to the right of the center. In some cases, the given center can be found within a band. Assuming again 32 equal bands (bands 0-31) and pixel values of 0-255, band 12 has a range of values from 96 to 103. If, for example, the SAO 74 filter determines a center of 99, then the center is within the range of pixel values covered by band 12. In this case, the SAO 74 filter can be configured to always include band 12 as one of the eight bands to the left of the center or always include band 12 as one of the eight bands to the right of the center. If the SAO filter 74 identifies band 12 as one of the eight bands to the left of the center, then the first group of bands includes bands 5-20, while the second group includes bands 0-4 and 21-31. If the SAO filter 74 identifies band 12 as one of the eight bands to the right of the center, then the first group of bands includes bands 6-21, while the second group includes bands 0-5 and bands 22- 31. The SAO filter 74 can also be configured to determine whether band 12 should be included with the bands to the left of the center or the bands to the right of the center, based on where it is within the band. Since 99 is closer to 96 than 103, the ODS filter 74 can be configured to identify band 12 as one of the bands to the left of the center in response to the determination of a center of 99. In response to the determination of a center 101, however, which is closer to 103 than 96, the SAO filter 74 can be configured to identify band 12 as one of the bands to the right of center.

[73] Although the techniques of this disclosure have been generally described using examples with thirty-two bands and two groups of sixteen bands, the techniques of this disclosure are generally applicable to ODS schemes that use more or less bands and more or less groups As one of many possible examples, the techniques of this disclosure can be used to determine a first group of four bands and a second group of twenty-eight bands, assuming that all displacements for the second group have a value of zero. In another example, sixteen bands, sixty-four bands or some other number of bands can be used.

[74] The SAO 74 filter can generate and signal displacements for bands based on these groupings. For each band group (again assuming sixteen band groups), sixteen offset values can be determined and signaled in the bit stream of encoded video for use in a video decoder. In one example, the number of bits required to send a group of sixteen band offset values can be reduced by taking advantage of the correlations between band offset values within the group. In another example, the band offset values of a neighboring partition can be used to predict the band offset values for a group of a current partition. The neighboring partition can be a causal partition for the current partition (that is, the neighboring partition has already been encoded). Examples of neighboring partitions include spatially neighboring partitions such as the previous or left partitions of the current partition, or temporarily neighboring partitions such as partitions in previously encoded frames (eg, co-located partitions).

[75] The techniques for finding a center can be applied to each LCU or to LCU groups. The techniques can, for example, be applied to regions of a frame that are determined analogously to the quadruple tree division discussed above. A plot can be divided into four regions; Each region can be divided into four smaller regions, and so on. The SAO filter 74 can determine core values as described above based on all pixel values within an LCU group or other frame region or it can determine core values by subsampling pixels in a way that uses less than all available pixel values. in the LCU or LCU group.

[76] The video encoder 20 of FIG. 4 represents an example of a video encoder configured to determine a central value for a set of pixels based on the values of; divide bands of pixel values into groups based on the central value; and, determine offset values for bands based on groups. The video encoder 20 can determine a center for a first set of pixels by determining an average value for the set of pixels, determining a median value for the set of pixels and / or determining the pixel where the peak of a histogram is located. The video encoder 20 can determine offset values for the bands based on the groups by generating a band offset value for a band of a first group based on a second band offset of the first group. The video encoder 20 can determine offset values for the bands based on the groups by generating a band offset value for a band of a first group based on a particular band offset for a contiguous partition. The video encoder 20 can also signal band offset values to a video decoder based on the groups. In some implementations, video encoder 20 can generate a syntactic element for inclusion in an encoded bit stream indicating the determined core value. In some cases, the syntactic element may

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

be an explicit indication of the given center for a group of bands, but the syntactic element does not necessarily have to be an explicit indication of the given center. For example, the central value may identify the first or last band of a particular group to allow a video decoder to reconstruct the same band groupings determined by video encoder 20. In this case, the band groupings reconstructed by the Video decoders can be aligned to the same center as the groupings determined by the video encoder, even if the center is not explicitly signaled in the encoded bit stream.

[77] FIG. 5 is a block diagram illustrating an example of video decoder 30 that can implement the techniques described in this disclosure. In the example of FIG. 5, the video decoder 30 includes an entropy decoding module 80, a prediction module 81, a reverse quantization module 86, a reverse transformation module 88, an adder 90, and reference image memory 92. The module Prediction 81 includes a motion compensation module 82 and an intra-prediction module 84. In some examples, the video decoder 30 can perform a decoding pass that is generally reciprocal to the described encoding pass with respect to the encoder of video 20 of FIG. Four.

[78] During the decoding process, the video decoder 30 receives a bit stream of encoded video representing video blocks of an encoded video fragment and associated syntactic elements, from the video encoder 20. The decoding module by entropy 80 of video decoder 30 performs decoding by entropy of the bit stream to generate quantified coefficients, motion vectors and other syntactic elements. The entropy decoding module 80 refers the motion vectors and other syntactic elements to the prediction module 81. The video decoder 30 can receive the syntactic elements at the level of the video fragment and / or the level of the video block.

[79] When the video fragment is encoded as an intracoded fragment (I), the intra prediction module 84 of the prediction module 81 can generate prediction data for a video block of the current video fragment, based on a mode of Intra-pointed prediction and previously decoded block data of the current frame or image. When the video frame is encoded as an inter-encoded fragment (ie, B, P or GPB), the motion compensation module 82 of the prediction module 81 generates predictive blocks for a video block of the current video fragment, based on motion vectors and other syntactic elements 55 received from the entropy decoding module 80. Predictive blocks can be generated from one of the reference images within one of the reference image lists. The video decoder 30 may construct the reference frame lists, List 0 and List 1, using predetermined construction techniques based on the reference images stored in the reference image memory 92.

[80] Motion compensation module 82 determines the prediction information for a video block of the current video fragment by analyzing the motion vectors and other syntactic elements, and uses the prediction information to generate the predictive blocks for video block current that is being decoded. For example, motion compensation module 82 uses some of the received syntactic elements to determine a prediction mode (for example, intra-prediction or inter-prediction), used to encode the video blocks of the video fragment, a type inter-prediction fragment (eg, fragment B, fragment P or fragment GPB), construction information for one or more of the reference image lists for the fragment, motion vectors for each block of inter-encoded video of the fragment, inter-prediction status for each block of inter-encoded video of the fragment and other information, to decode the video blocks in the current video fragment.

[81] Motion compensation module 82 can also perform interpolation based on interpolation filters. Motion compensation module 82 may use interpolation filters such as those used by video encoder 20 during video block encoding to calculate interpolated values for sub-integer pixels of reference blocks. In this case, the motion compensation module 82 can determine the interpolation filters used by the video encoder 20 from the received syntactic elements and use the interpolation filters to generate predictive blocks.

[82] The inverse quantization module 86 quantifies inversely, that is, it quantifies, the quantified transformation coefficients provided in the bit stream and decoded by the entropy decoding module 80. The inverse quantization process may include the use of a quantification parameter calculated by the video encoder 20 for each video block in the video fragment, to determine a degree of quantification and, also, a degree of inverse quantification that should be applied. The inverse transformation module 88 applies an inverse transformation, for example, an inverse DCT, an entire inverse transformation or an inverse transformation process conceptually similar to the transformation coefficients, in order to produce residual blocks in the pixel domain.

[83] After the motion compensation module 82 generates the predictive block for the current video block based on the motion vectors and other syntactic elements, the video decoder 30 forms a decoded video block by adding the residual blocks of the reverse transformation module

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

88 and the corresponding predictive blocks generated by the motion compensation module 82. Adder 90 represents the component or components that perform this summing operation. The decoded video blocks formed by adder 90 can then be filtered by an unblocking filter 93, an ODS filter 94 and an adaptive loop filter 95. The decoded video blocks in a given frame or image are then stored in reference image memory 92, which stores reference images used for subsequent motion compensation. Reference image memory 92 also stores decoded video for later display in a display device, such as display device 32 of FIG. one.

[84] The SAO 94 filter can be configured to apply the same filtering (for example, edge offset and band offset) as the SAO 74 filter described above. Thus, as part of the realization of the band offset, the SAO filter 74 can determine a central value for a first group of bands as described above in reference to FIG. 4. The central value can, for example, be determined based on a metric such as the average pixel value, the median pixel value, or determined based on a pixel where the peak of a histogram corresponds to an LCU or a group of LCU .

[85] In some implementations, the SAO 94 filter may receive an indication of a particular center in a video encoder in the video bit stream. The indication can take the form of a real central value calculated by a video encoder or it can take the form of an identification of a first band for the first group, an indication of a last band of a first part of a second group, a average value for the first group, or other information that allows the SAO 94 filter to determine the same band grouping used by a video encoder. Based on a first group and a second group determined based on the core value, the SAO filter 94 can generate offset values based on the information received in a bit stream of encoded video. For one of the groups, such as the second group, it can be assumed that all offset values are equal to zero.

[86] For an LCU group, the SAO 94 filter can determine a central value by determining an average pixel value for the LCU group. The SAO 94 filter can determine the average pixel value by adding the pixel values of all or some of the pixels in the LCU group and dividing by the number of pixels whose values are part of the sum, the resulting value being the average value of the pixel for the LCU group. The SAO 94 filter unit can then use the average pixel value determined in the manner described above to determine band groupings.

[87] Alternatively, for an LCU group, the SAO 94 filter can determine a central value by determining an average pixel value for the LCU group. The SAO 94 filter can determine the median pixel value by identifying the value at which half of the pixel values of the LCU group is greater than the value, and half of the pixel values for the LCU group is less than the value. The SAO filter unit 94 can then use the average pixel value determined in the manner described above to determine band groupings.

[88] Alternatively, for an LCU group, the SAO 94 filter can determine a central value by determining where the peak of a histogram for an LCU or an LCU group is located. In general, a histogram can be considered a statistical representation of the distribution of pixel values for the LCU group. Therefore, the SAO 94 filter can determine the peak of the histogram by identifying the pixel value or pixel values that appear most frequently in the LCU group. The SAO filter unit 94 can then use the determined value of the peak in the histogram in the manner described above to determine band groupings.

[89] The three examples of how to determine a central value (eg, mean, median, peak of the histogram) can be performed using one or both filters SAO 74 and SAO 94 and are not the only three possible techniques by which you can determine a central value. It is contemplated that the techniques of this disclosure, which generally refer to dynamically determining band groupings to indicate band offset values, may be compatible with other techniques of this type to determine a central value.

[90] The video decoder 30 of FIG. 5 represents an example of a video decoder configured to determine a central value for a set of pixels based on pixel values, divide the bands of pixel values into groups based on the central value and determine offset values for the bands based on the groups. In some implementations, video decoder 30 may determine the core value upon receiving a syntactic element indicating the core value. As described above, the syntactic element may or may not be an explicit indication of the core value. In some implementations, video decoder 30 can determine the central value by determining an average value for the set of pixels, determining an average value for the set of pixels, and / or determining the pixel where the peak of a histogram is located. The video decoder 30 can determine offset values for the bands based on the groups by generating a band offset value for a band of a first group based on a second band offset of the first group. The video decoder 30 can determine offset values for the bands based on the groups generating a value of

5

10

fifteen

twenty

25

30

35

40

Four. Five

fifty

55

60

65

band offset for a band of a first group based on a particular band offset for a contiguous partition. The video decoder 30 can also reconstruct the band offset values signaled from a video encoder based on the groups.

[91] Thus, the video decoder 30 of FIG. 5 represents an example of a video decoder configured to determine a central value for a set of pixels based on pixel values for the group; divide bands of pixel values into groups based on the central value; and determine offset values for bands based on groups. The video decoder 30 can determine the central value by receiving a syntactic element indicating the central value, determine the central value by determining an average value for the set of pixels, determine the central value by determining a median value for the set of pixels, or determine The central value comprises determining the pixel where the peak of a histogram is located. The pixel set may, for example, include a group of larger coding units (LCUs). The video decoder 30 can determine offset values for the bands based on the groups by generating a band offset value for a band of a first group based on a second band offset of the first group. The video decoder 30 can determine offset values for the bands based on the groups by generating a band offset value for a band of a first group based on a particular band offset for a contiguous partition.

[92] FIG. 6 shows a flow chart illustrating the SAO filtering techniques described in this disclosure. The techniques of FIG. 6 can be performed by a video encoder with an SAO filtering module such as a video encoder 20 or a video decoder 30. The techniques of FIG. 6 will be described with reference to video decoder 30, although it should be understood that the techniques of FIG. 6 are applicable to a wide range of video encoding devices and are not limited only to video decoders.

[93] Video decoder 30 determines a central value for a set of pixels based on pixel values (610). The video decoder 30 can, for example, determine the central value by determining an average value for the set of pixels, determining an average value for the set of pixels, determining a pixel where the peak of a histogram is located. The video decoder 30 divides the bands of pixel values into groups based on the central value (620). The video decoder 30 determines the offset values for the bands based on the groups (630). The video decoder 30 can determine offset values for the bands based on the groups by generating a band offset value for a band of a first group based on a second band offset of the first group. The video decoder 30 can also determine offset values for the bands based on the groups by generating a band offset value for a band of a first group based on a particular band offset for a contiguous partition.

[94] In one or more examples, the described functions can be implemented in hardware, software, firmware or any combination of these. If implemented in software, the functions can be stored in or transmitted through, as one or more instructions or code, in a computer-readable medium or executed through a hardware-based processing unit. Computer-readable media may include computer-readable storage media, which correspond to a tangible media such as data storage media or communication media that include any media that facilitates the transfer of a computer program from one location to another. , for example, according to a communication protocol. Thus, the computer-readable media may generally correspond to (1) a computer-readable tangible storage media that are non-transient, or (2) a communication medium such as a signal or a carrier wave. The data storage media may be any available media that can be accessed from one or more computers or one or more processors to retrieve instructions, code and / or data structures for the implementation of the techniques described in this disclosure. A computer program product may include a computer readable medium.

[95] By way of example, and not limitation, such computer-readable storage media may comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, flash memory or any other means that can be used to store desired program code in the form of instructions or data structures and which can be accessed by a computer. Also, any connection duly receives the name of computer readable media. For example, if instructions are transmitted from a website, server or other remote source using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL) or wireless technologies such as infrared, radio and microwave, then coaxial cable, fiber optic cable, twisted pair, DSL or wireless technologies such as infrared, radio and microwave are included in the definition of medium. However, it should be understood that computer readable storage media and data storage media do not include connections, carrier waves, signals or other transient media, but instead are geared towards tangible non-transient storage media. The term disc, as used herein, includes a compact disc (CD), a laser disc, an optical disc, a digital versatile disc (DVD), a flexible disc and a Blu-ray disc, where some discs they usually emit data magnetically, while other discs emit data optically with lasers. Combinations of the above should also be included within the scope of computer readable media.

[96] The instructions can be executed by one or more processors, such as one or more digital signal processors (DSP), general purpose microprocessors, application specific integrated circuits (ASIC), field programmable logic formations (FPGA) ) or other integrated or discrete logic circuits

5 equivalents Accordingly, the term "processor", as used herein, may refer to any of the above structures or to any other structure suitable for the implementation of the techniques described herein. In addition, in some aspects, the functionality described herein may be provided within dedicated hardware and / or software modules configured for encoding and decoding, or incorporated into a combined codec. In addition, the techniques could

10 be fully implemented in one or more circuits or logical elements.

[97] The techniques of the present disclosure can be implemented in a wide variety of devices or devices, including a cordless telephone, an integrated circuit (IC) or a set of ICs (eg, a chipset). Various components, modules or units are described in this disclosure to emphasize aspects

15 functional devices configured to perform the disclosed techniques, but do not necessarily require their realization through different hardware units. Instead, as described above, various units may be combined in a codec hardware unit or provided by means of a group of interoperable hardware units, which include one or more processors such as those described above, in conjunction with software and / or adequate firmware.

twenty

[98] Various examples have been described. These and other examples are within the scope of the following claims.

Claims (12)

10 fifteen twenty 2. 25 3. 30 Four. 35 40 5. 45 6. 50 7. 55 8. 9. 60 A video coding method for signaling, in a coded bit stream, the offset values used in the band-based adaptive sample shift (ODS) filtering, the method comprising: grouping bands of pixel values into two or more band groups, in which each band of the pixel value bands comprises a range of pixel values, and in which each group of the two or more band groups comprises a or more bands; determine offset values for bands based on the two or more band groups; characterized by determine (610) core values for each of a plurality of sets of pixels, in which for each set the core value is based on the actual values of the pixels in the set; Y wherein, for each pixel set, the grouping (620) of the pixel bands into two or more band groups is determined dynamically based on the central value for the set so that the grouping of bands into pixel sets is adaptive The method according to claim 1, wherein the method is performed by a video decoder and for each set of pixels, in which determining the central value comprises receiving a syntactic element, and in which dividing the pixel value bands in two or more groups of bands it comprises determining the two or more groups of bands based on the syntactic element, the method further comprising the application of an adaptive sample displacement filter, ODS, based on the offset values. The method according to claim 1, wherein the method is performed by a video encoder, and wherein the method further comprises for each set of pixels: generate a syntactic element for inclusion in an encoded bit stream, in which the syntactic element indicates the central value. The method according to claim 1, wherein the determination of the core value comprises one of: a) determine an average value for the set of pixels; or b) determine a median value for the set of pixels; or c) determine the value where the peak of a histogram of the set pixel values is located. The method according to claim 1, wherein each set of pixels comprises a group of larger coding units, LCU. The method according to claim 1, wherein determining offset values for bands based on the two or more band groups comprises generating a band offset value for a band of a first group based on a second band offset from the first group. The method according to claim 1, wherein determining offset values for the bands based on the two or more groups of bands comprises generating a band offset value for a band of a first group in the current set of pixels based on a determined band offset for another set of pixels. A computer readable storage medium that stores operable instructions to cause one or more processors to carry out the procedure of any one of claims 1 to 7. A video coding apparatus (20; 30) for signaling, in a coded bit stream, the offset values used in the band-based adaptive sample shift (ODS) filtering, the apparatus comprising: means (74; 94) for grouping bands of pixel values into two or more groups of bands, in which each band of the pixel value bands comprises a range of pixel values and in which each group of the two or more groups of bands comprises one or more bands; means (74; 94) for determining offset values for bands based on the two or more 5 10 fifteen twenty 25 30 35 40 Four. Five band groups; characterized by: means (74; 94) for determining the core values for each of a plurality of sets of pixels, in which for each set the core value is based on the actual values of the pixels in the set; Y wherein the means for grouping bands of pixel values into two or more band groups are dynamically configured to determine the division of pixel bands into two or more band groups based on the core value for the set so that the Grouping bands into pixel sets is adaptive. 10. The apparatus according to claim 9, wherein the apparatus comprises a video decoder (30), and wherein the means for determining the central value comprise means for receiving a syntactic element and means for dividing the value bands Pixel in two or more band groups comprise means for determining the two or more band groups based on the syntactic element; and wherein the apparatus further comprises means for applying an adaptive sample displacement filter, ODS, based on the displacement values. 11. The apparatus of claim 9, wherein the apparatus comprises a video encoder (20), and wherein the apparatus further comprises: means for generating a syntactic element for inclusion in an encoded bit stream, in which the syntactic element indicates the central value. 12. The apparatus of claim 9, wherein the means for determining the core value comprises one of: a) means for determining an average value for the set of pixels; or b) means for determining an average value for the set of pixels; or c) means to determine the value where the peak of a histogram of the set pixel values is located. 13. The apparatus of claim 9, wherein each set of pixels comprises a group of larger coding units, LCU. 14. The apparatus of claim 9, wherein the means for determining the offset values for the bands based on the two or more groups of bands comprise means for generating a value of band offset for a band of a first group based on a second band offset of the first group. 15. The apparatus according to claim 9, wherein the means for determining the offset values for the bands based on the two or more band groups comprise means for generating a band offset value for a band of a first group in the current set of pixels based on a particular band offset for another set of pixels.